SketchGraphs is a data set of 15 million sketches from CAD models which researchers say can help make predictions that help humans creating CAD designs.
Machine learning performed by neural networks is a popular approach to developing artificial intelligence, as researchers aim to replicate brain functionalities for a variety of applications.
A paper in the journal Applied Physics Reviews, by AIP Publishing, proposes a new approach to perform computations required by a neural network, using light instead of electricity. In this approach, a photonic tensor core performs multiplications of matrices in parallel, improving speed and efficiency of current deep learning paradigms.
In machine learning, neural networks are trained to learn to perform unsupervised decision and classification on unseen data. Once a neural network is trained on data, it can produce an inference to recognize and classify objects and patterns and find a signature within the data.
However, to dismiss the subject as fantastical or unnecessary would be akin to telling scientists 100 years ago that landing on the moon was also irrelevant.
This is because, for pioneers and champions of artificial intelligence, quantum computing is the holy grail. It’s not a make-believe fantasy; rather, it’s a tangible area of science that will take our probability-driven world into a whole new dimension.
The core of GPT-3, which is a creation of OpenAI, an artificial intelligence company based in San Francisco, is a general language model designed to perform autofill. It is trained on uncategorized internet writings, and basically guesses what text ought to come next from any starting point. That may sound unglamorous, but a language model built for guessing with 175 billion parameters — 10 times more than previous competitors — is surprisingly powerful.
With attention focused on a pandemic and an election, AI has taken a major leap forward.
Performance of photon-based neural network processor is 100-times higher than electrical processor.
Once Upon a Time I Lived on Mars: Space, Exploration, and Life on Earth by Kate Greene St. Martin’s Press, 2020 hardcover, 240 pp. ISBN 978−1−250−15947−2 US$27.
While the robotic missions launching to Mars this year have a wide range of science goals, they are widely seen as precursors for eventual human missions to the Red Planet. NASA’s Mars 2020 mission includes an experiment called MOXIE that will demonstrate a way to produce oxygen from the carbon dioxide in the Martian atmosphere, a capability that will be essential for future human expeditions. NASA’s fiscal year 2021 budget proposal included a request to start work on a Mars Ice Mapper mission, an orbiter that would search for subsurface ice deposits that could be resources for future human expeditions.
Much of the planning for future Mars missions is focused on various capabilities needed to safely transport humans to the surface of Mars and bring them back. But beyond technologies like in situ resource utilization and supersonic retropropulsion are more mundane, but no less essential, matters: How will the crew eat? How will they deal with boredom on the long mission? How will they get along with one another in a confined space?
Need a robot with a soft touch? A team of Michigan State University engineers has designed and developed a novel humanoid hand that may be able to help.
In industrial settings, robots often are used for tasks that require repetitive grasping and manipulation of objects. The end of a robot where a human hand would be found is known as an end effector or gripper.
“The novel humanoid hand design is a soft-hard hybrid flexible gripper. It can generate larger grasping force than a traditional pure soft hand, and simultaneously be more stable for accurate manipulation than other counterparts used for heavier objects,” said lead author Changyong Cao, director of the Laboratory for Soft Machines and Electronics at MSU and assistant professor in Packaging, Mechanical Engineering, and Electrical and Computer Engineering.
Researchers at Tel Aviv University, led by Prof. Yaniv Assaf of the School of Neurobiology, Biochemistry and Biophysics and the Sagol School of Neuroscience and Prof. Yossi Yovel of the School of Zoology, the Sagol School of Neuroscience, and the Steinhardt Museum of Natural History, conducted a first-of-its-kind study designed to investigate brain connectivity in 130 mammalian species. The intriguing results, contradicting widespread conjectures, revealed that brain connectivity levels are equal in all mammals, including humans.
“We discovered that brain connectivity —namely the efficiency of information transfer through the neural network —does not depend on either the size or structure of any specific brain,” says Prof. Assaf. “In other words, the brains of all mammals, from tiny mice through humans to large bulls and dolphins, exhibit equal connectivity, and information travels with the same efficiency within them. We also found that the brain preserves this balance via a special compensation mechanism: when connectivity between the hemispheres is high, connectivity within each hemisphere is relatively low, and vice versa.”
Participants included researchers from the Kimron Veterinary Institute in Beit Dagan, the School of Computer Science at TAU and the Technion’s Faculty of Medicine. The paper was published in Nature Neuroscience on June 8.
Carnegie Mellon today showed off new research into the world of robotic navigation. With help from the team at Facebook AI Research (FAIR), the university has designed a semantic navigation that helps robots navigate around by recognizing familiar objects.
The SemExp system, which beat out Samsung to take first place in a recent Habitat ObjectNav Challenge, utilizes machine learning to train the system to recognize objects. That goes beyond simple superficial traits, however. In the example given by CMU, the robot is able to distinguish an end table from a kitchen table, and thus extrapolate in which room it’s located. That should be more straightforward, however, with a fridge, which is both pretty distinct and is largely restricted to a singe room.
